Search

We propose a compact and simple all-optical subcarrier-multiplexed (SCM) label-swapping system employing an integrated electroabsorption modulation laser and a semiconductor optical amplifier based Mach-Zehnder interferometer wavelength converter. The experiments demonstrated error-free all-optical label swapping for the 155-Mb/s label and 10-Gb/s payload over two optical label-switching network nodes with less than 0.7-dB power penalty on the payload. The majority. of the components in this SCM label-swapping system can be realized on a semiconductor platform (e.g., InP), which implies a step toward possible monolithic or hybrid integration of the all-optical label-swapping system in the future.

In this study, a new type of ceramic composite (PZSC/ZnO) was successfully synthesized using N, P, and S co‐doped carbon microspheres derived from polyphosphazene with zinc oxide. The morphology and structure were characterized by Fourier transform infrared spectroscopy, X‐ray photoelectron spectroscopy, scanning electron microscopy, energy‐dispersive spectroscopy, Raman, X‐ray diffraction, BET, and thermogravimetric. Batch experiments were used to explore the adsorption performance under different pH values, initial concentrations, contact times, and temperatures. The results showed that the maximum adsorption capacity of the PZSC/ZnO ceramic composite for uranium was 470.5 mg/g (pH = 5.5, t = 40 min, and T = 25°C). The adsorption process followed the nonlinear Langmuir model and the pseudo‐second‐order kinetic model, demonstrating that the monolayered combination of U(VI) with the PZSC/ZnO ceramic composite and the adsorption mechanism was chemical adsorption. Thermodynamic data revealed that the adsorption was a spontaneous endothermic process. Furthermore, P and N may be involved in the adsorption of uranyl ions through binding to the O atoms from ZnO. The main forces between U(VI) and the PZSC/ZnO ceramic composite were attributed to ZnO and heteroatoms doped in the material. [ABSTRACT FROM AUTHOR]

As a promising technology, Flex Ethernet (FlexE) helps to realize deterministic and ultra-low latency in metro and transport networks. Meanwhile, previous studies have confirmed the advantages of the symbiosis of FlexE and elastic optical network (EON) (i.e., a FlexE-over-EON) on resource utilization and cost-effectiveness. In this paper, we consider the cross-layer restoration (CLR) in FlexE-over-EONs based on the FlexE-aware architecture. Specifically, we address the situation where an outage happened on one FlexE switch in the packet layer to bring it offline temporarily and then the affected client flows need to be recovered quickly and proactively. Three CLR strategies are first proposed to fully explore the flexibility of FlexE-over-EON for restoring the affected flows. Then, with the strategies, we formulate an integer linear programming (ILP) model and design an auxiliary graph (AG) based algorithm to reroute the affected flows as well as minimize the additional operational expense (OPEX) incurred during the CLR. Extensive simulations verify the effectiveness of our proposed CLR algorithms. [ABSTRACT FROM AUTHOR]

The fast development of multi-layer packet-over-optical networks has made network monitoring and troubleshooting increasingly complicated. This has stimulated people to combine machine learning (ML) and software-defined networking (SDN) to realize multi-layer network automation. Despite its initial successes, the vulnerabilities of multi-layer network automation have not been fully explored. This work studies how to mislead the ML-based classifiers for anomaly detection. Specifically, we design two adversarial-sample-based attack schemes based on the white-box attack (WBA) and black-box attack (BBA) strategies, respectively, to eavesdrop and tamper legitimate telemetry data samples and generate adversarial samples adaptively, for disturbing ML-based classifiers and in turn misleading network automation to make incorrect decisions. Compared with WBA, BBA makes the attack scheme more practical by minimizing the dependency on pre-knowledge of the target ML-based classifiers. Considering different types of ML-based classifiers, we build a real-world packet-over-optical testbed and leverage the telemetry samples collected in it to demonstrate that our proposed BBA scheme can interact with the network quietly to train itself, generate well-crafted adversarial samples to tamper legitimate telemetry samples in the hard-to-detect way, and mislead ML-based classifiers in the network automation system to severely affect their performance on anomaly detection. [ABSTRACT FROM AUTHOR]

Recently, IP over elastic optical network (IP-over-EON) has become a promising architecture for metro and core networks. This work studies how to visualize both layers of an IP-over-EON in real time, at different granularities (e.g., at flow-level, lightpath-level, and link-level), and with self-adaptivity. Specifically, we consider the multilayer application of in-band telemetry (INT) and propose entropy-driven adaptive INT (namely, EntropyINT). We introduce stateful processing to programmable data plane (PDP) switches for EntropyINT, such that they can make local decisions to determine whether and what type of telemetry data about the IP and EON layers should be encoded in each packet. The local decisions are designed to be based on the amount of information that can be conveyed by telemetry data to the network automation system. Meanwhile, we make EntropyINT cooperate with out-of-band monitoring, to detect and locate exceptions in the EON layer. Our proposal is implemented in a real-world testbed of IP-over-EON, to evaluate its assistance to network automation. Experimental results verify the effectiveness of our proposal, and indicate that the telemetry data collected by EntropyINT and out-of-band monitoring can better assist the machine learning in network automation, for status prediction and anomaly detection. [ABSTRACT FROM AUTHOR]

Hybrid optical/electrical datacenter network (HOE-DCN) uses the inter-rack networks that consist of both electrical Ethernet switches and optical cross-connects (OXCs), for better cost-efficiency and scalability. Meanwhile, to provision dynamic network services well, the operator of an HOE-DCN needs to deploy virtual networks (VNTs) and remap them adaptively. Therefore, this work studies the problem of VNT remapping in an HOE-DCN from a novel perspective, i.e., the remapping schemes should be optimized for not only the network status after the remapping but also the transition to realize it. Specifically, we model this problem as a bilevel optimization, where the upper-level optimization aims at selecting proper virtual machines (VMs) to migrate such that the estimated latency of VM migration can be minimized, and the lower-level optimization determines the actual scheme of VNT remapping for minimizing the number of resource hot-spots. We first formulate a bilevel mixed integer linear programming (BMILP) model for the bilevel optimization, and then propose a polynomial time algorithm based on enumeration to solve it approximately. Extensive simulations verify the effectiveness of our proposal. [ABSTRACT FROM AUTHOR]

Network function virtualization (NFV) in elastic optical datacenter interconnections (EO-DCIs) enables flexible and timely deployment of network services. However, as the service provisioning of virtual network function service chains (vNF-SCs) in an EO-DCI needs to orchestrate the allocations of IT resources in datacenters (DCs) and spectrum resources on fiber links dynamically, it is a complex and challenging problem. In this work, we model the problem as a Markov decision process (MDP), and propose a hierarchical deep reinforcement learning (DRL) model based on graph neural network (GNN), namely, HRLOrch, to tackle it. To ensure its universality and scalability, we design the policy neural network (NN) in HRLOrch based on a GNN. As the GNN-based policy NN can operate on the graph-structured network state of an EO-DCI directly, it can adapt to an arbitrary EO-DCI topology without any structural changes. Then, through analysis, we find that the EO-DCI is a sparse reward environment if we want to train a DRL model to minimize the blocking probability of vNF-SCs in it directly. To address this issue, we design a hierarchical DRL with lower-level and upper-level models to improve the convergence performance of training. Specifically, we make the lower-level DRL optimize the provisioning scheme of each vNF-SC to minimize its resource usage, while the upper-level one coordinates the provisioning of all the active vNF-SCs to minimize the overall blocking probability. Hence, the lower-level and upper-level DRL models operate cooperatively in the training to optimize the dynamic provisioning of vNF-SCs. Our simulations demonstrate the universality and scalability of HRLOrch, and confirm that it can outperform the existing algorithms for vNF-SC provisioning in an EO-DCI. [ABSTRACT FROM AUTHOR]

The service provisioning in multi-domain software-defined elastic optical networks (SD-EONs) is an interesting but difficult problem to tackle, because the basic problem of lightpath provisioning, i.e., the routing and spectrum assignment (RSA), is $\mathcal {NP}$ -hard, and each domain is owned and operated by a different carrier. Therefore, even though numerous RSA heuristics have been proposed, there does not exist a universal winner that can always achieve the lowest blocking probability in all the scenarios of a multi-domain SD-EON. This motivates us to revisit the inter-domain provisioning problem in this paper by leveraging deep reinforcement learning (DRL). Specifically, we propose DeepCoop, which is an inter-domain service framework that uses multiple cooperative DRL agents to achieve scalable network automation in a multi-domain SD-EON. DeepCoop employs a DRL agent in each domain to optimize intra-domain service provisioning, while a domain-level path computation element (PCE) is introduced to obtain the sequence of the domains to go through for each lightpath request. By sharing a restricted amount of information among each other, the DRL agents can make their decisions distributedly. To ensure scalability and universality, we design the action space of each DRL agent based on well-known RSA heuristics, and architect the agents based on the soft actor-critic (SAC) scenario. We run extensive simulations to evaluate DeepCoop, and the results show that DeepCoop can adapt to the dynamic environment in a multi-domain SD-EON to always select the best RSA heuristic for minimizing blocking probability, and it outperforms the existing algorithms on inter-domain provisioning in various scenarios. Moreover, we verify that the distributed training implemented in DeepCoop ensures its universality and scalability (i.e., its training and operation do not depend on the topology of the SD-EON). [ABSTRACT FROM AUTHOR]

The fast development of high-performance and flexible SmartNICs and programmable data plane switches (PDP-SWs) has motivated people to consider the deployment of virtual network functions (vNFs) on them. Hence, together with traditional virtual machines (VMs), SmartNICs and PDP-SWs form heterogeneous network function virtualization (NFV) platforms for realizing vNF service chains (vNF-SCs). In this work, we consider the transition from software-based homogeneous NFV platforms to the heterogeneous ones, and study how to optimize the service upgrade of vNF-SCs. Specifically, the service upgrade is divided into two steps, which are 1) selecting servers/switches in the substrate network (SNT) to upgrade, which is done by adding SmartNICs to servers and replacing traditional switches with PDP-SWs, under a fixed budget, and 2) redeploying the existing vNF-SCs in the updated SNT to maximize the quality-of-service (QoS) improvement on latency reductions. We first formulate an integer linear programming (ILP) model to optimize the overall service upgrade, then design two correlated optimizations for its two steps, and finally propose polynomial-time approximation algorithms to solve the optimizations. The results of extensive simulations confirm that our proposed algorithm outperforms the existing benchmarks in various network scenarios, and achieves better tradeoff between performance and time-efficiency. [ABSTRACT FROM AUTHOR]

This Special Issue contains a collection of outstanding papers covering several recent advances in optical networks for 5G communications and beyond. Papers are organized into four categories: network resource planning; optical access networks; optical fronthaul solutions; and autonomous and data-driven network management. In this introduction, a brief overview of the field is given, followed by a summary of the seventeen papers of this Special Issue, and a discussion of future directions in the field. [ABSTRACT FROM AUTHOR]

To closely monitor the performance of lightpaths in an elastic optical network (EON), people need to rely on real-time and fine-grained spectrum monitoring. This, however, will generate tremendous telemetry data, which can put great pressure on both the control and data planes. In this work, we design and experimentally demonstrate AutoSpecheck, which is a DL-assisted network automation (DLaNA) system that can realize highly-efficient and automatic spectrum inspection for anomaly detection in EONs. Specifically, we architect AutoSpecheck based on the software-defined EON (SD-EON) architecture, and propose techniques to greatly reduce the loads of data reporting (in the data plane) and data analyzing (in the control plane). To reduce the loads of data reporting, we leverage the AutoEncoder (AE) technique to design a spectrum data compression method. To improve the efficiency of data analytics, we first design a coarse filtering module (CFM) to let the control plane filter out most of the normal data before invoking the DL-based anomaly detection. Then, to address the difficulty of labeling massive spectrum data, we develop a DL-based anomaly detection based on semi-supervised learning. Our experimental demonstrations consider two representative intra-channel anomalies (i.e., the filter drifting and in-band jamming), and the results confirm that AutoSpecheck can achieve highly-efficient and automatic spectrum inspection for anomaly detection in EONs. [ABSTRACT FROM AUTHOR]

Recently, the programmable data plane (PDP) switches have been considered as the key enablers for in-network computing. However, the limited memory resources in them for flow tables might restrict their performance. This work addresses this challenge by studying how to optimize the placements of flow tables in the external memory on multiple servers, and to access them with remote direct memory access (RDMA) for ensuring low latency. Specifically, we consider a data-center network (DCN) that uses PDP switches as top-of-rack (ToR) switches, and propose and optimize the hybrid flow table installation (hFT-INST) on each ToR switch. With hFT-INST, the switch can either store flow tables in its local memory or use RDMA to install and access them remotely in its rack servers. We first design the protocol and operation procedure of hFT-INST. Then, regarding the key problem of hFT-INST, i.e., how to place the flow tables on the external memory on different servers, we take a few practical parameters into account, and formulate a mixed integer linear programming (MILP) model to tackle it. Next, the optimization in the MILP is transformed into a capacitated facility location problem (CFLP) with additional constraints. We further transform it into a ${k}$ -median problem through pre-processing, and design a polynomial-time approximation algorithm to solve the problem. Extensive simulations confirm the performance of our proposed algorithm. We also prototype our design of the hFT-INST, and conduct experiments to demonstrate its feasibility. [ABSTRACT FROM AUTHOR]

This article discusses DeepMDR, which is a deep learning (DL)-assisted control plane (CP) system to realize scalable and protocol-independent path computation in multi-domain packet networks. We develop DeepMDR based on ONOS, make it support protocol-oblivious forwarding (POF) in the data plane, facilitate a hierarchical CP architecture for multi-domain operations, and propose a DL model to achieve fast and high-quality path computation in each domain. Simulation results verify that our DL-assisted routing module achieves better trade-off between path computation time and routing performance than existing approaches. The effectiveness of our proposed DeepMDR is also demonstrated with experiments, which show that it serves inter-domain flow requests quickly with a processing capacity of ∼166,000 messages/s or higher. [ABSTRACT FROM AUTHOR]

Hybrid optical/electrical datacenter networks (HOE-DCNs) build inter-rack networks with both electrical Ethernet switches and optical cross-connects (OXCs), and have been considered as a promising DCN architecture. However, to adapt to the dynamic network environment, the reconfiguration of virtual networks (VNTs) in an HOE-DCN still faces the unique difficulty that the HOE-DCN's topology can change because of the one-to-one connectivity of OXCs. To the best of our knowledge, this problem still has not been fully explored. In this article, we address this problem, and consider how to achieve effective VNT reconfiguration in an HOE-DCN such that the IT resource usages in racks can be re-balanced with the migration of virtual machines (VMs). We first formulate a mixed integer linear programming (MILP) to describe the VNT reconfiguration. Then, we solve the problem with two steps, 1) obtaining the VM migration schemes to balance the loads on racks and 2) determining the reconfiguration schemes of related virtual links (VLs) and the OXC. For the first step, we propose a polynomial-time approximation algorithm by leveraging linear relaxation. Then, we tackle the optimization of the second step by developing an algorithm that involves a linear-time dynamic programming and an integer linear programming (ILP). To solve the ILP time-efficiently, we propose another polynomial-time approximation algorithm based on Lagrangian relaxation. Our simulations confirm the effectiveness of the proposed approximation algorithms and verify that the overall procedure including them outperforms the existing approach. [ABSTRACT FROM AUTHOR]

With the evolution of Internet infrastructure and network services, multilayer in-band network telemetry (ML-INT) and data analytics (DA) have been considered as key enabling techniques to realize real-time and fine-grained network monitoring, especially for backbone IP-over-Optical networks. However, the existing ML-INT&DA systems have privacy and security issues, because plaintext ML-INT data is reported from the data plane and gets analyzed in the control plane. In this work, we address these issues by designing a privacy-preserving ML-INT&DA system for IP-over-Optical networks. We first leverage vector homomorphic encryption (VHE) to design a lightweight encryption scheme, which overcomes the security breaches due to eavesdropping and preserves the delicate correlations buried in multi-dimensional ML-INT data. Then, we develop an effective data compression scheme to further encode the encrypted ML-INT data and make the results suitable for hash-based signature. The signature is for data certification and enables the DA in the control plane to verify the integrity of received ML-INT data. Hence, the threats from data tampering are removed. Next, we architect a deep learning (DL) model that can directly operate on encrypted ML-INT data for anomaly detection. Finally, we implement the proposed ML-INT&DA system, and experimentally demonstrate its effectiveness in a real IP over elastic optical network (IP-over-EON) testbed, whose key elements, i.e., optical line system (OLS), bandwidth-variable wavelength-selective switches (BV-WSS’) and programmable data plane (PDP) switches, are all commercial products. [ABSTRACT FROM AUTHOR]

Recently, the combination of artificial intelligence (AI) and software-defined networking (SDN) has attracted intensive research interests because it realizes and promotes AI-assisted network automation (AIaNA). Despite the initial successes of AIaNA, its vulnerabilities, i.e., the downside of the reduction of human involvement achieved by it, have not been carefully explored. In this work, we use software-defined IP over elastic optical networks (SD-IPoEONs) as the background, and study how to mislead the AIaNA system in them. Specifically, we target our attack on the deep neural network (DNN) based traffic predictor in the AIaNA system, and design an adversarial module (ADVM) that can craft and inject adversarial traffic samples adaptively to disturb its operation. We consider two approaches to design the ADVM, i.e., the deep reinforcement learning (DRL) based on deep deterministic policy gradient (DDPG), and the generative adversarial network (GAN) model. Our proposed ADVM can monitor and interact with a dynamic SD-IPoEON to train itself on-the-fly. This enables it to generate and inject adversarial samples in the most disturbing and hard-to-detect way and to severely affect the AIaNA's performance on multilayer service provisioning. Specifically, IP flows will be served incorrectly to result in unnecessary congestions/under-utilizations on lightpaths, and erroneous network reconfigurations will be invoked frequently. Simulation results confirm the effectiveness of our ADVM designs, and show that the GAN-based ADVM achieves better attack effects with smaller perturbation strength. [ABSTRACT FROM AUTHOR]

With the dense deployment of millimeter wave (mmWave) front ends and popularization of bandwidth-intensive applications, shared backhaul in fiber-wireless (FiWi) networks is still facing a bandwidth crunch. To alleviate the backhaul pressure, in this paper, caching capability is enabled at the edge of FiWi networks, i.e., optical network unit access points (ONU-APs). On the other hand, as both power budget and backhaul bandwidth in FiWi networks are constrained, it is challenging to properly leverage power for caching and that for wireless transmission to achieve superior system performance. As caching has a significant impact on resource allocation, we reconsider performance optimization and analysis in cache-enabled FiWi networks. Firstly, in the cache-enabled FiWi network with mmWave, we formulate the joint power allocation and caching problem, with the goal to maximize the downlink throughput. A two-stage algorithm is then proposed to solve the problem. Secondly, to investigate the theoretical capacity of the cache-enabled FiWi network with mmWave, we derive an upper bound of the downlink throughput by analyzing properties of the average rate of wireless links. Particularly, we show that appropriate power allocation for wireless transmission and caching at ONU-APs is essential to achieve higher throughput. The numerical and simulation results validate our theoretical analysis and demonstrate the proposed algorithm can approach the analytical upper bound. [ABSTRACT FROM AUTHOR]

It is known that by leveraging programmable data plane, in-band network telemetry (INT) can be realized to provide a powerful and promising method to collect realtime network statistics for monitoring and troubleshooting. However, existing INT implementations still exhibit a few drawbacks such as lack of runtime-programmability and relatively high overheads due to per-packet operation. In this work, we propose and design a runtime-programmable selective INT system, namely, Sel-INT, to resolve these issues. Specifically, we first design a runtime-programmable selective INT scheme based on protocol oblivious forwarding (POF), and then prototype our design by extending the famous OpenvSwitch (OVS) platform to obtain a software switch that supports Sel-INT and implementing a Data Analyzer to parse, extract and analyze the INT data. Our implementation of Sel-INT is verified and evaluated in a real network testbed that consists of a few stand-alone software switches. The experimental results demonstrate that Sel-INT can not only adjust the sampling rate of INT in runtime but also program the corresponding data types dynamically, and they also confirm that our proposal can ensure proper accuracy and timeliness for network monitoring while greatly reducing the overheads of INT. [ABSTRACT FROM AUTHOR]

To better provision fast-emerging network applications with various quality-of-service demands, datacenter network (DCN) operators need an effective network orchestration scheme that can coordinate IT and bandwidth resources for differentiated services in a timely manner. In this work, we consider a hybrid optical–electrical DCN (HOE-DCN) and study how to achieve scalable knowledge-defined network orchestration (KD-NO) for managing the delay-sensitive and delay-tolerant applications in it. For delay-sensitive applications, we leverage a multi-agent scheme to distribute the tasks of placing virtual machines (VMs) in server racks and routing VM traffic in electrical–optical inter-rack clouds to two cooperative deep reinforcement learning modules, respectively. Then, we utilize a classic-algorithm-based module to provision delay-tolerant applications with the residual resources in the HOE-DCN. We design the operation and coordination procedure of the KD-NO system and build a small HOE-DCN testbed that consists of four server racks to demonstrate its performance experimentally. Experimental results indicate that our KD-NO system can make timely and correct network orchestration decisions and have better convergence performance compared with the existing benchmark. [ABSTRACT FROM AUTHOR]

Recently, the fast development of backbone networks has made the traffic, services, and infrastructure of packet-over-optical networks increasingly complicated. This stimulates research and development on fine-grained and real-time performance monitoring and troubleshooting. In this article, we propose a ProML-INT system to oversee packet-over-optical networks in real time and enable customized performance monitoring and troubleshooting. We introduce the system design in detail, and explain how to control the overhead of multilayer INT ML-INT by inserting INT fields in packets selectively. Experiments demonstrate the ProML-INT system, a small-scale packet-over-optical network testbed. The experimental results confirm that our proposal can monitor packet and optical layers jointly in real time, and the homemade data analyzer in it can leverage artificial intelligence to identify the root causes of exceptions in packet-over-optical networks correctly and promptly. [ABSTRACT FROM AUTHOR]

In this article, we discuss the design and implementation of a novel DCN system, which utilizes a knowledge-defined NO-M to operate a HOEDCN cost-effectively and energy-efficiently. The motivations behind the proposed HOE-DCN system are the urgent need to address the scalability, energy, and manageability issues in existing DCN systems. To realize the knowledge-defined NO-M, we follow the principle of predictive analytics in the human brain to design three artificial intelligence modules based on deep learning and make them operate collaboratively. The proposed HOE-DCN system is implemented in a network testbed, and we conduct experiments that involve both control and data plane operations to demonstrate its advantages. The experimental results show that the HOE-DCN simultaneously achieves high-performance service provisioning and improved energy efficiency. Furthermore, by analyzing the pros and cons of the HOE-DCN system, we also point out several directions to work on in the future. [ABSTRACT FROM AUTHOR]

For datacenter networks (DCNs), it is always important to have an effective network orchestration scheme that can coordinate the usages of IT and bandwidth resources timely. In this paper, we consider the hybrid optical/electrical DCNs (HOE-DCNs) and propose a knowledge-defined network orchestration (KD-NO) system for them. The KD-NO system follows the predictive analytics in human behaviors, which includes forecasting based on memory and decision making based on knowledge. To explain the design of our KD-NO system, we first discuss how to fetch low-level knowledge from the telemetry data about the resource utilization in an HOE-DCN. Then, we describe how to optimize the HOE-DCN's configuration for network orchestration. Specifically, we design an online scheme based on deep reinforcement learning, and make sure that it can extract high-level knowledge from the low-level input and come up with optimal HOE-DCN configurations on-the-fly. We prototype the proposed KD-NO system and demonstrate it in an HOE-DCN testbed. The experiments run Hadoop applications in the testbed and show that our KD-NO system can make timely and correct decisions in different experimental schemes by leveraging the two-level knowledge, maintain a high matching degree between the HOE-DCN's configuration and the applications running in it, and thus, effectively reduce the job completion time. [ABSTRACT FROM AUTHOR]

Recently, with the development of IP and elastic optical networks (EONs), the network control and management (NC&M) scheme for IP-over-EONs, which can facilitate effective cross-layer orchestration (XLyr-O), has become an interesting but challenging research topic. In this paper, we consider a software-defined IP-over-EON (SD-IPoEON), leverage deep learning (DL) to analyze and predict the traffic fluctuation on established lightpaths in it, and design a proactive DL-assisted XLyr-O scheme. Specifically, we study the DL-assisted XLyr-O scheme from algorithm design to system prototype. A DL module based on the long/short-term memory based neural network (LSTM-NN) is first designed and optimized for precise IP traffic prediction. Then, we develop algorithms to explore the traffic prediction for realizing proactive XLyr-O to deal with hard/soft failures constantly, i.e., making intelligent online decisions to re-groom and reroute IP flows and to reconfigure lightpaths such that the performance tradeoff among lightpath utilization, congestion probability, and reconfiguration frequency is balanced well. Finally, we implement our proposed algorithm in a small-scale but real SD-IPoEON testbed to prototype the DL-assisted XLyr-O, and conduct experiments with it. Experimental results demonstrate that compared with the reactive benchmark without DL-assistance, our proposal not only invokes less network reconfigurations but also reduces packet losses significantly. [ABSTRACT FROM AUTHOR]

This paper proposes DeepRMSA, a deep reinforcement learning framework for routing, modulation and spectrum assignment (RMSA) in elastic optical networks (EONs). DeepRMSA learns the correct online RMSA policies by parameterizing the policies with deep neural networks (DNNs) that can sense complex EON states. The DNNs are trained with experiences of dynamic lightpath provisioning. We first modify the asynchronous advantage actor-critic algorithm and present an episode-based training mechanism for DeepRMSA, namely, DeepRMSA-EP. DeepRMSA-EP divides the dynamic provisioning process into multiple episodes (each containing the servicing of a fixed number of lightpath requests) and performs training by the end of each episode. The optimization target of DeepRMSA-EP at each step of servicing a request is to maximize the cumulative reward within the rest of the episode. Thus, we obviate the need for estimating the rewards related to unknown future states. To overcome the instability issue in the training of DeepRMSA-EP due to the oscillations of cumulative rewards, we further propose a window-based flexible training mechanism, i.e., DeepRMSA-FLX. DeepRMSA-FLX attempts to smooth out the oscillations by defining the optimization scope at each step as a sliding window, and ensuring that the cumulative rewards always include rewards from a fixed number of requests. Evaluations with the two sample topologies show that DeepRMSA-FLX can effectively stabilize the training while achieving blocking probability reductions of more than 20.3% and 14.3%, when compared with the baselines. [ABSTRACT FROM AUTHOR]

Recently, a number of WiFi access points (APs) have been densely deployed to provide widely available, high-performance Internet services. As such, an energy efficiency issue becomes crucial toward the design of green wireless local area networks (WLANs). In this paper, we propose a resource and re-association scheduling algorithm (referred to RAS) based on Benders’ decomposition to reduce the energy consumption. In particular, we endeavor to aggregate WLAN users on the small number of APs and turn off many APs without compromising users’ quality of experience (QoE) and system coverage. We conduct the analysis by using real trace data and formulate the energy minimization as the mixed integer nonlinear programming (MINLP) problem. We then transform and solve the original problem through the RAS algorithm. For practical implementation, we further propose the fast RAS (Fast-RAS) algorithm to relax the binary integer constraints and transform the MINLP problem into the nonlinear programming (NLP) problem. The relaxed problem then can be solved by using Feasible Pump algorithm with the reduced computational complexity. We evaluate the performance of RAS and Fast-RAS algorithms via extensive simulations. The results demonstrate that the Fast-RAS algorithm can achieve up to 20% improvement of energy saving comparing with existed methods. [ABSTRACT FROM AUTHOR]

The combination of network virtualization and software-defined networking enables an infrastructure provider to create software-defined virtual networks (vSDNs) over a shared substrate network (SNT), for supporting new network services more timely and cost-effectively. Meanwhile, as both the services and traffic in the Internet are becoming more and more dynamic, how to properly maintain vSDNs in a dynamic network environment exhibits increasing importance but still has not been fully explored. In this paper, we conduct a study on how to realize proactive and hitless vSDN reconfiguration to balance the utilization of ternary content-addressable memory (TCAM) in a dynamic SNT. Specifically, we consider both algorithm design and system prototyping. From the algorithmic perspective, we try to solve the problems of “what to reconfigure” and “how to reconfigure”. A selection algorithm is designed to proactively choose the virtual switches (vSWs) that should be migrated to other substrate switches for balancing TCAM utilization, i.e., solving what to reconfigure. Then, for the problem of how to reconfigure, i.e., where to re-map the selected vSWs and the virtual links connecting to them, we formulate a mixed integer linear programming model to solve it exactly, and design two heuristics to improve time efficiency. Next, we move to the system part, implement the proposed algorithms in our protocol-oblivious forwarding enabled network virtualization hypervisor system, and conduct experiments to demonstrate proactive and hitless vSDN reconfiguration. The experimental results indicate that our proposal does make vSDN reconfiguration transparent to the vSDNs’ virtual controllers and proactive, and when reconfiguring a vSDN with live traffic, it achieves hitless operations without traffic disruption. [ABSTRACT FROM AUTHOR]

This paper proposes a self-taught anomaly detection framework for optical networks. The proposed framework makes use of a hybrid unsupervised and supervised machine learning scheme. First, it employs an unsupervised data clustering module (DCM) to analyze the patterns of monitoring data. The DCM enables a self-learning capability that eliminates the requirement of prior knowledge of abnormal network behaviors and therefore can potentially detect unforeseen anomalies. Second, we introduce a self-taught mechanism that transfers the patterns learned by the DCM to a supervised data regression and classification module (DRCM). The DRCM, whose complexity is mainly related to the scale of the applied supervised learning model, can potentially facilitate more scalable and time-efficient online anomaly detection by avoiding excessively traversing the original dataset. We designed the DCM and DRCM based on the density-based clustering algorithm and the deep neural network structure, respectively. Evaluations with experimental data from two use cases (i.e., single-point detection and end-to-end detection) demonstrate that up to $99\%$ anomaly detection accuracy can be achieved with a false positive rate below $1\%$. [ABSTRACT FROM AUTHOR]

In this paper, we propose an incentive-driven virtual network function service chaining (VNF-SC) framework for optimizing the cross-stratum resource provisioning in multi-broker orchestrated inter-datacenter elastic optical networks (IDC-EONs). The proposed framework employs a non-cooperative hierarchical game-theoretic mechanism, where the resource brokers and the VNF-SC users play the leader and the follower games, respectively. In the leader game, the brokers calculate VNF-SC service schemes for users and compete for the provisioning tasks. While in the follower game, the users compete for VNF-SC services for jointly optimizing the resource cost and the received quality-of-service. We first elaborate on the modeling of the follower game, discuss the existence of Nash equilibrium and propose a mixed-strategy gaming approach enabled by an auxiliary graph-based algorithm to facilitate users selecting the most appropriate service schemes. Then, under the assumption that the brokers are aware of the principle of the follower game, we present the model for the leader game and develop a time-efficient heuristic algorithm for brokers to compete for the provisioning tasks. Simulations show that the proposed incentive-driven VNF-SC framework significantly improves the network throughput (i.e., $>4.8 \times$ blocking reduction) while assisting users and brokers in achieving higher utilities compared with existing solutions. [ABSTRACT FROM AUTHOR]

With flexibility in optical layer, elastic optical network (EON) has been considered as a competitive candidate to architect next-generation backbone networks. Routing and spectrum assignment (RSA) is a key problem for the service provisioning in EONs. The RSA problem is $\mathcal {NP}$ -hard even in elastic optical rings. Numerous heuristics have been proposed, and they can generally be categorized into two types: Route-First (RF) and Spectrum-First (SF). Although most previous work demonstrated by numerical simulations that the SF algorithms always outperform the RF ones, there is a lack of theoretical analysis on the reasons causing the performance difference between the two types of RSA algorithms. In this paper, we aim at proposing a unified theoretical framework for the performance analysis of RSA algorithms by leveraging conflict graphs, which offers a new perception on the optimality of RSA algorithms. To validate the proposed framework, we apply it in elastic optical rings (with cycle topology), and theoretically analyze the number of edges of the conflict graphs for RF and SF algorithms. Different from the literature, we obtain an interesting observation that neither the RF nor the SF can surpass the other in elastic optical rings under different traffic distributions, and their performances have a strong correlation to the edge count of their conflict graph. This observation provides a new perspective, i.e., conflict graph, to explore the property of RSA algorithms. [ABSTRACT FROM AUTHOR]

This paper demonstrates, for the first time to our knowledge, hierarchical learning framework for inter-domain service provisioning in software-defined elastic optical networking (EON). By using a broker-based hierarchical architecture, the broker collaborates with the domain managers to realize efficient global service provisioning without violating the privacy constrains of each domain. In the proposed hierarchical learning scheme, machine learning-based cognition agents exist in the domain managers as well as in the broker. The proposed system is experimentally demonstrated on a two-domain seven-node EON testbed for with real-time optical performance monitors (OPMs). By using over 42000 datasets collected from OPM units, the cognition agents can be trained to accurately infer the Q-factor of an unestablished or established lightpath, enabling an impairment-aware end-to-end service provisioning with an prediction Q-factor deviation less than 0.6 dB. [ABSTRACT FROM AUTHOR]

In multi-domain elastic optical networks with alien wavelengths, each domain needs to consider intradomain and interdomain alien traffic to estimate and guarantee the required quality of transmission (QoT) for each lightpath and perform provisioning operations. This paper experimentally demonstrates an alien wavelength performance monitoring technique and machine-learning-aided QoT estimation for lightpath provisioning of intradomain/ interdomain traffic. Testbed experiments demonstrate modulation format recognition, QoT monitoring, and cognitive routing for a 160 Gbaud alien multi-wavelength light- path. By using experimental training datasets from the testbed and an artificial neural network, we demonstrated an accurate optical-signal-to-noise ratio prediction with an accuracy of ~95% when using 1200 data points. [ABSTRACT FROM AUTHOR]

To realize cost-effective and adaptive network control and management (NC&M) on inter-datacenter optical networks (IDCONs), people have considered network virtualization to let the operator of an IDCON work as an infrastructure provider (InP), which can create virtual optical networks (VONs) over the IDCON for tenants. In this paper, we use this network scenario as the background, and try to integrate deep learning (DL) based traffic prediction in the NC&M of the IDCON and the VONs created over it. We first design the service provisioning framework in which each tenant uses a DL module to predict the traffic in its VON and will submit a VON reconfiguration request to the InP, when it sees a significant mismatch between future traffic and the allocated resources in its VON. Then, the InP will invoke the VON reconfiguration to make the VON be better prepared for future traffic. An adaptive and scalable DL-based traffic predictor is proposed together with a cognitive service provisioning algorithm to exploit the temporal and spatial characteristics of interDC traffic and achieve effective service provisioning based on precise and timely traffic prediction. Next, we consider the situation where a tenant leverages “machine-learning-as-a-service” and outsources the training of its DL module to a third-party entity for overcoming its resource limitations, and analyze the induced vulnerabilities due to data poisoning. Our simulation results indicate that with our proposal, the InP can invoke VON reconfigurations timely and improve the service provisioning performance of each VON significantly. Meanwhile, the results also demonstrate that our data poisoning scheme can easily bypass the normal validation of the DL module and generate significant adversarial effects. [ABSTRACT FROM AUTHOR]

This work addresses the relatively long setup latency and complicated network control and management caused by on-demand virtual network function service chain (vNF-SC) provisioning in inter-datacenter elastic optical networks. We first design a provisioning framework with resource pre-deployment to resolve the aforementioned challenge. Specifically, the framework is designed as a discrete-time system, in which the operations are performed periodically in fixed time slots (TS). Each TS includes a pre-deployment phase followed by a provisioning phase. In the pre-deployment phase, a deep-learning (DL) model is designed to predict future vNF-SC requests, then lightpath establishment and vNF deployment are performed accordingly to pre-deploy resources for the predicted requests. Then, the system proceeds to the provisioning phase, which collects dynamic vNF-SC requests from clients and serves them in real-time by steering their traffic through the required vNFs in sequence. In order to forecast the high-dimensional data of future vNF-SC requests accurately, we design our DL model based on the long/short-term memory-based neural network and develop an effective training scheme for it. Then, the provisioning framework and DL model are optimized from several perspectives. We evaluate our proposed framework with simulations that leverage real traffic traces. The results indicate that our DL model achieves higher request prediction accuracy and lower blocking probability than two benchmarks that also predict vNF-SC requests and follow the principle of the proposed provisioning framework. [ABSTRACT FROM AUTHOR]

Inter-datacenter (inter-DC) networks are essential for large enterprises to deliver high-quality services to end-users. Since DCs are vulnerable to natural disasters, an inter-DC network operator needs an effective emergency backup plan to evacuate the endangered data out in case of a progressive disaster whose status can be predicted by an early warning system. In this paper, we try to solve the problem of emergency backup in inter-DC networks with progressive disasters. We first utilize the time-expanded network (TEN) approach to model the time-variant inter-DC network during a progressive disaster as a variant TEN (VTEN) and convert the dynamic flow scheduling for emergency backup to a static one. Then, with the VTEN, we formulate an optimization model to maximize the profit from the emergency backup in consideration of data values and resource costs. Although this large-scale optimization can be solved in a distributed way by leveraging the alternation direction method of multipliers (ADMM), we find that one of its subproblems is nontrivial in the distributed setting. We propose a novel inexact ADMM approach to resolve the issue induced by the subproblem, and prove that the proposed algorithm can converge to the optimal solution. The results from extensive simulations confirm that our algorithm is robust and time-efficient, and outperforms several benchmarks in terms of backup profit and running time. [ABSTRACT FROM AUTHOR]

Software-defined networking (SDN) has been considered as a break-through technology for the next-generation Internet. It enables fine-grained flow control that can make networks more flexible and programmable. However, this might lead to scalability issues due to the possible flow state explosion in SDN switches. SDN-based source routing can reduce the volume of flow-tables significantly by encoding the path information into packet headers. In this paper, we leverage the protocol-oblivious forwarding instruction set to design protocol-oblivious source routing (POSR), which is a protocol-independent, bandwidth-efficient, and flow-table-saving packet forwarding technique. We lay out the packet format for POSR, come up with the packet processing pipelines for realizing unicast, multicast, and link failure recovery, and implement POSR in a protocol-oblivious forwarding-enabled SDN network system. Experiments are then performed in a network testbed, which consists of 14 stand-alone SDN switches, to validate the advantages of POSR. Specifically, we compare POSR with several OpenFlow-based benchmarks for unicast, multicast, and link failure recovery, and confirm that POSR can reduce flow-table utilization effectively, shorten path setup latency and expedite link failure recovery. [ABSTRACT FROM PUBLISHER]

This paper investigates the problem of how to optimize the provisioning of virtual network function service chains (VNF-SCs) in elastic optical inter-datacenter networks (EO-IDCNs) under elastic optical networking and DC capacity constraints. We take advantage of the broker-based hierarchical control paradigm for the orchestration of cross-stratum resources and propose to realize incentive-driven VNF-SC provisioning with a noncooperative mixed-strategy gaming approach. The proposed gaming model enables tenants to compete for VNF-SC provisioning services due to revenue and quality-of-service incentives and therefore can motivate more reasonable selections of provisioning schemes. We detail the modeling of the game, discuss the existence of the Nash equilibrium states, and design an auxiliary graph-based heuristic algorithm for tenants to compute approximate equilibrium solutions in the games. A dynamic resource pricing strategy, which can set the prices of network resources in real time according to the actual network status, is also introduced for EO-IDCNs as a complementary method to the game-theoretic approach. Results from extensive simulations that consider both static network planning and dynamic service provisioning scenarios indicate that the proposed game-theoretic approach facilitates both higher tenant and network-wide profits and improves the network throughput as well compared with the baseline algorithms, while the dynamic pricing strategy can further reduce the request blocking probability with a factor of ?2.4?. [ABSTRACT FROM PUBLISHER]

Application-driven networks (ADNs) aim to build logically separate virtual networks (VNs) to meet the distinct demands of different applications. In this paper, we study how to realize the on-demand slicing of application-driven virtual software-defined elastic optical networks (vSD-EONs) based on the concept of ADN. We design the network system for on-demand application-driven vSD-EON slicing and demonstrate the building and operating of application-driven vSD-EONs with it experimentally. Specifically, our experimental demonstrations consider three scenarios: (1) the tenant's application requires high availability for the data plane, (2) the tenant's application is interactive and thus requires short end-to-end latency, and (3) the tenant's application needs an enhanced physical-layer security guarantee. With an experimental testbed that consists of commercial optical transmission facilities, bandwidth-variable wavelength-selective switches, erbium-doped fiber amplifiers, and high-performance servers with both optical and electrical ports, we verify that the proposed vSD-EON slicing system can build vSD-EONs on-demand according to tenants' application demands and operate them correctly. [ABSTRACT FROM PUBLISHER]

Due to the flexibility and adaptivity of elastic optical networks (EONs), IP-over-EON would be a promising infrastructure for next-generation backbone networks. As a backbone network usually carries massive traffic, one always needs to properly address its network survivability issue. The network survivability of an IP-over-EON can be affected not only by the fiber cuts in the EON layer but also by the router outages in the IP layer. In this work, we study how to realize the cross-layer orchestration to address IP router outages with cost-efficient multilayer restoration (MLR) in IP-over-EONs. Specifically, we consider the situation in which a single router outage happens in an IP-over-EON and propose MLR algorithms to minimize the additional operational expense (OPEX) due to MLR. We first design three MLR strategies to fully explore the flexibility and adaptivity of IP-over-EONs. Then, with the strategies, we formulate an integer linear programming (ILP) model to find the MLR scheme in which the additional OPEX due to incremental usages of sliceable bandwidth-variable transponders (SBV-Ts) and frequency slots (FSs) and lightpath reconfigurations is minimized. We also propose an auxiliary graph (AG) based heuristic algorithm to reduce the time complexity. The proposed algorithms are evaluated with extensive simulations, and the results indicate that, compared with an existing benchmark, they can effectively reduce the additional OPEX of MLR. [ABSTRACT FROM PUBLISHER]

This letter tries to optimize the availability-aware service provisioning (AaSP) with failure-independent path-protecting pre-configured cycles (FIPP- p -cycles) in elastic optical networks (EONs). We propose a novel AaSP-FIPP scheme by leveraging bandwidth-squeezed restoration, develop a mathematical model to analyze the service availability of the scheme, and design a topology partitioning method to improve its scalability. [ABSTRACT FROM PUBLISHER]

It is known that software-defined elastic optical networks (SD-EONs) facilitate optical networking that provides better network programmability, more powerful manageability, and more flexible service provisioning capability. Moreover, the hierarchical architecture of multibroker-based multidomain SD-EONs cannot only improve the network scalability but also maintain the autonomy of each administrative domain. In this paper, we study why and how the brokers should cooperate with each other to provision interdomain lightpaths in multibroker-based multidomain SD-EONs. We first formulate a cooperative market in which the brokers negotiate about their market shares (i.e., the opportunities to provision interdomain lightpaths) and seek for a mutual agreement with Nash bargaining [1]. Then, we design a mathematical model to describe the market as well as the brokers’ behaviors in it. An effective algorithm is derived from the model to solve the Nash bargaining problem for allocating lightpath requests among the brokers. The proposed algorithm also addresses the resource collision during request provisioning and can achieve collision-free request allocation. Extensive simulations verify the effectiveness of our proposal. [ABSTRACT FROM PUBLISHER]

Network function virtualization (NFV) is a promising technology to decouple the network functions from dedicated hardware elements, leading to the significant cost reduction in network service provisioning. As more and more users are trying to access their services wherever and whenever, we expect the NFV-related service function chains (SFCs) to be dynamic and adaptive, i.e., they can be readjusted to adapt to the service requests’ dynamics for better user experience. In this paper, we study how to optimize SFC deployment and readjustment in the dynamic situation. Specifically, we try to jointly optimize the deployment of new users’ SFCs and the readjustment of in-service users’ SFCs while considering the trade-off between resource consumption and operational overhead. We first formulate an integer linear programming (ILP) model to solve the problem exactly. Then, to reduce the time complexity, we design a column generation (CG) model for the optimization. Simulation results show that the proposed CG-based algorithm can approximate the performance of the ILP and outperform an existing benchmark in terms of the profit from service provisioning. [ABSTRACT FROM PUBLISHER]

In elastic optical networks, two lightpaths sharing common fiber links might have to be isolated in the spectrum domain with a proper guard-band to prevent crosstalk and/or reduce physical-layer security threats. Meanwhile, the actual requirements on guard-band sizes can vary for different lightpath pairs, because of various reasons. Therefore, in this paper, we consider the situation in which the actual guard-band requirements for different lightpath pairs are different, and formulate the distance spectrum assignment (DSA) problem to investigate how to assign the spectrum resources efficiently in such a situation. We first define the DSA problem formally and prove its \mathcal NP -hardness and inapproximability. Then, we analyze and provide the upper and lower bounds for the optimal solution of DSA, and prove that they are tight. In order to solve the DSA problem time-efficiently, we develop a two-phase algorithm. In its first phase, we obtain an initial solution and then the second phase improves the quality of the initial solution with random optimization. We prove that the proposed two-phase algorithm can get the optimal solution in bipartite DSA conflict graphs and can ensure an approximate ratio of \mathcal O(\log (|V|)) in complete DSA conflict graphs, where $|V|$ is the number of vertices in the conflict graph, i.e., the number of lightpaths to be considered. Numerical results demonstrate our proposed algorithm can find near-optimal solutions for DSA in various conflict graphs. [ABSTRACT FROM AUTHOR]